CD47 has emerged as a novel therapeutic target in hematologic cancers. Leukemias expressing the highest levels of CD47 have the worst prognosis. Increased CD47 expression is thought to protect cancer cells from phagocytic clearance by sending a don't eat me signal to macrophages via SIRPa, an inhibitory receptor that prevents phagocytosis of CD47-bearing cells. Anti-CD47 mAbs (CD47mAbs) that block the CD47/SIRPa interaction enhance phagocytosis of cancer cells and contribute to control of tumor burden in human to mouse xenograft leukemia models (). However, there are mechanisms by which CD47 mAbs can attack transformed cells that have not yet been exploited in the war on cancer. A particular anti-human CD47mAb (clone 1F7) has a direct, cytotoxic effect on human leukemias. MAb 1F7 kills CD47-bearing tumor cells without the action of complement or other immune cells. Instead, mAb 1F7 acts via a non-apoptotic mechanism that involves a direct CD47-dependent attack on mitochondria, discharging their membrane potential (DYm) and destroying the ATP-generating capacity of the cell leading to rapid cell death. It is noteworthy that mAb 1F7 does not kill normal leukocytes, which also express CD47, but only those cells that are activated by transformation. Thus normal circulating cells, all of which express CD47, are spared while cancer cells are selectively killed by the tumor-toxic CD47mAb. Importantly, mAb 1F7 also blocks binding of SIRPa to CD47 and thus it can act via two mechanisms: (1) direct tumor cytotoxicity and (2) promoting phagocytosis of the dead and dying tumor cells. We hypothesize that a single mAb that can accomplish both functions will be superior to one that only blocks CD47/SIRPa binding. In this phase I project, we will characterize CD47mAbs with tumor-toxic activity toward mouse leukemia cells and then test these mAbs for efficacy in treating mouse acute promyelocytic leukemias (APL). In phase II of this program, we will humanize our candidate CD47mAb and evaluate its efficacy and safety. The PHASE I SPECIFIC AIMS are: 1. Identification/characterization of a tumor-toxic CD47mAb selected from a panel of CD47 mAbs (series 400 mAbs) that react with CD47 from many species including mouse and human. 2. Tumor-toxic CD47mAbs identified in Aim 1 will be tested for efficacy in syngeneic mouse APL models in vivo. APL cells will be tagged with luciferase and green fluorescent protein (GFP) for sensitive detection. We will compare anti-cancer efficacy of the tumor-toxic CD47mAbs to a CD47mAb that simply blocks CD47/SIRPa binding. This study will also give us a preliminary indication of in vivo safety/tolerability of these CD47mAbs that can bind CD47 on both the leukemia and the host's normal cells, a situation reflecting human cancer therapy.